Unrestricted class structure compatibility and focused throughput assessment for your model's performance evaluation
Python 3.8+
pip install model-meter
---> 100%- Create a file
main.pywith:
from model_meter import ModelMeter
from ultralytics import YOLO
# Load the model
model = YOLO()
# Create the meter
meter = ModelMeter(model)
# Run the meter
throughput, avg_time_per_call = meter.measure(
params=('./contents/dog.jpg',),
method_name="__call__",
)
print(f'Throughput: {throughput} images/second')
print(f'Avg time per call: {avg_time_per_call} seconds')Run the script with:
$ python main.py
...
Throughput: 14.794271662158552 images/second
Avg time per call: 0.0675937297107937 secondsThis project is licensed under the terms of the MIT license.
